Prevalence and correlates of severe anxiety among front-line nurses during and after the COVID-19 pandemic: a large-scale multi-center study

This study is a large-scale multicenter investigation aimed at examining the incidence of severe anxiety and related factors among frontline nurses in China during and after the COVID-19 pandemic. The cross-sectional design was chosen because it allows for the simultaneous assessment of anxiety levels, along with various demographic, job-related, and psychological factors, during two distinct time periods: during the pandemic and after its peak. This approach enables us to identify trends and relationships that may inform public health policies and mental health interventions tailored for nurses. The study design followed the STROBE guidelines for observational studies.

We collected data on demographic characteristics, sleep disorders, depression, and anxiety using the “Wenjuanxing” online survey platform. Prior to the survey, informed consent was obtained from each participant. This study adhered to the ethical standards outlined in the 1964 Helsinki Declaration and its subsequent amendments. The study protocol was approved by the Institutional Review Board (IRB) of Tianjin Mental Health Center (Approval No: 2023-028).

The study was conducted in 27 provinces of China from November to December 2022 and from April to July 2023, recruiting frontline nurses. These nurses were engaged in direct patient care during the COVID-19 pandemic, working in hospitals, clinics, and other healthcare facilities. The diverse geographical representation ensures that the findings reflect a broad range of experiences among frontline healthcare workers, taking into account the varying healthcare infrastructure and regional differences in response to the pandemic. Unlike other countries, China was still in the midst of the COVID-19 pandemic from November to December 2022. April to July 2023 marked the period after the peak of the COVID-19 pandemic.

The sample size was determined based on an estimation of the prevalence of severe anxiety in similar populations, with an expected prevalence of 50% during the pandemic and 10% afterward, based on prior research [4, 5]. To achieve a confidence level of 95% with a margin of error of 5%, the minimum required sample size was calculated to be 384 per group. However, to increase statistical power and account for potential non-responses or incomplete data, we aimed to recruit approximately 1,200 nurses during the pandemic and 1,200 post-pandemics. Ultimately, 1,153 nurses were recruited during the pandemic, and 816 completed the survey. Post-pandemic, we followed up with 763 nurses from the original cohort. The inclusion criteria were frontline nurses who worked directly with COVID-19 patients and voluntarily participated. The exclusion criteria included nurses without formal nursing qualifications, those absent from work due to pregnancy or medical leave, and those who refused to participate.

We employed a convenience sampling method across China, recruiting participants through online platforms, which allowed us to reach a wide geographic distribution of nurses. Nurses who were unable to complete the survey online were excluded from the analysis, ensuring data consistency and reducing potential bias. The study flowchart is shown in Fig. 1.

We utilized a self-designed general demographic questionnaire, covering nurses’ demographic characteristics, job characteristics, and perceptions of the impact of the COVID-19 pandemic on their lives and work. For example, the definition of weekly exercise frequency was: exercising for more than 30 min per session was considered as one instance of exercise; 0 indicated never exercising, 1–2 times per week, and 3 times or more per week. High economic pressure and high work pressure were primarily based on nurses’ self-reports. Additionally, we investigated whether nurses felt that COVID-19 disrupted their lives and whether they were worried that clinical work could lead to COVID-19 infection. This instrument was validated through expert review and pilot testing prior to the survey distribution. It was designed to capture a broad range of variables that could potentially influence mental health outcomes.

PSQI was used to assess sleep disorders [25]. PSQI comprises 19 questions divided into seven components: sleep duration, sleep disturbances, sleep onset latency, subjective sleep quality, sleep efficiency, daytime dysfunction, and use of sleep medications. Each component has a score ranging from 0 to 3, with the total PSQI score ranging from 0 to 21. A total PSQI score ≥ 8 indicates sleep disorders [26]. We selected this scale because it is widely used in clinical and research settings, providing a standardized method to assess sleep disturbances, which are commonly reported in healthcare workers, especially in the context of the COVID-19 pandemic. Anxiety was measured using the Zung Self-Rating Anxiety Scale (SAS) [27], and depression was measured using the Self-Rating Depression Scale (SDS) [28]. The cutoff values for SAS and SDS were 50 [29] and 40 [30], respectively. Severe anxiety was defined as a standard SAS score ≥ 70 [29]. The SAS was used to measure anxiety due to its ease of administration and well-documented psychometric properties. The SAS is a commonly used tool to assess the severity of anxiety symptoms in clinical and non-clinical populations. Similarly, the SDS was selected to evaluate depression due to its simplicity and strong psychometric properties. The SDS is effective in assessing the severity of depressive symptoms in diverse populations, including healthcare workers. In this study, the SAS demonstrated a Cronbach’s alpha of 0.89, and the SDS demonstrated a Cronbach’s alpha of 0.91, indicating strong internal reliability. In addition, the Cronbach’s alpha was 0.86, confirming its robustness in evaluating sleep disorders in this cohort.

To examine the factors influencing severe anxiety, binary logistic regression analysis was conducted. Severe anxiety (SAS ≥ 70) was treated as the dependent variable, and factors showing significant differences in the chi-square test or t-test (e.g., nursing title, night shift frequency, depression, and sleep disorders) were included as independent variables. Odds ratios (ORs) and 95% confidence intervals (CIs) were reported for each factor. Receiver Operating Characteristic (ROC) analysis was performed using GraphPad Prism 9.0 to evaluate the predictive value of relevant variables for severe anxiety.

A total of 816 nurses working during the COVID-19 pandemic and 763 nurses recruited after the pandemic were included. The results revealed that the incidence of severe anxiety among nurses during the COVID-19 pandemic was significantly higher than that after the pandemic (52.3% vs. 8.0%, P < 0.001). As shown in Table 1, significant differences were observed between nurses during and after the COVID-19 pandemic in terms of nursing title, monthly night shift frequency, educational level, weekly exercise frequency, COVID-19 infection status, severe economic pressure, severe work pressure, and life disruption (P < 0.01). For example, after the COVID-19 pandemic, there was a significant decrease in the frequency of night shifts, exercise frequency, and COVID-19 infection rate among nurses (P < 0.01). However, the proportion of nurses who believed that COVID-19 disrupted their lives significantly increased (52.0% vs. 78.6%, P < 0.001).

The comparison of SAS, SDS, and PSQI scores among nurses during and after the COVID-19 pandemic is presented in Table 2. The SAS and SDS scores of nurses after the COVID-19 pandemic were significantly lower than those during the pandemic (P < 0.001). Furthermore, the total PSQI score, subjective sleep quality, sleep duration, severe anxiety, and sleep medication scores of nurses after the COVID-19 pandemic were significantly lower than those during the pandemic (P < 0.01). After Bonferroni correction, significant differences still existed in SAS scores, SDS scores, total PSQI scores, sleep duration, severe anxiety, and sleep medication scores (P Bonferroni < 0.05/10 = 0.005).

To further assess the relationship between severe anxiety (SAS ≥ 70) and factors with significant statistical differences, binary logistic regression analysis was conducted. As shown in Table 3, binary logistic regression analysis was performed on the covariates of severe anxiety. The results showed that during the COVID-19 pandemic, nurses with more than 10-night shifts per month had a 1.520 times higher risk of severe anxiety compared to those with less than 5-night shifts per month (OR = 1.520, 95% CI 1.074–2.152). Interestingly, nurses infected with COVID-19 had a 0.711 times lower risk of severe anxiety compared to those not infected (OR = 0.711, 95% CI 0.528–0.958).

After the COVID-19 pandemic, the risk of severe anxiety among nurses in supervisory positions and above was 0.440 times lower than that of nurses below supervisory positions (OR = 0.440, 95% CI 0.211–0.916). Additionally, depression (OR = 1.108, 95% CI 1.067–1.152) and sleep disorders (OR = 1.193, 95% CI 1.101–1.292) were independent risk factors for severe anxiety among nurses. ROC analysis showed that SAS scores and PSQI scores had good predictive value for severe anxiety among nurses after the pandemic (area under the ROC curve for SDS score was 0.798, and for PSQI score was 0.799, respectively) (Fig. 2).